metadata
dataset_info:
features:
- name: item_name
dtype: string
- name: audio
dtype: audio
- name: text
dtype: string
splits:
- name: train
num_bytes: 83465834725
num_examples: 250000
download_size: 70286415823
dataset_size: 83465834725
configs:
- config_name: default
data_files:
- split: train
path: data/train-*